Full text: ARCH+ : Studienhefte für architekturbezogene Umweltforschung und -planung (1969, Jg. 2, H. 5-8)

development of a simulation model. Each stage moves 
one step further from the real world, thus introducing 
the possibility of interpretive and judgmental error. 
Simulation model building starts with the real world 
situation, and proceeds through identification of an area 
of analysis and establishment of model criteria to speci- 
fication of the outputs required for model corroboration 
and decision-making. 
The first abstraction is from reality to a general theory of 
how the world operates. This is the kind of intuitional 
knowledge urban planners have developed about the 
economic, physical and social relationships of cities. It 
tends to be vague, qualitative, comprehensive and full 
of uncertainties; yet it is usually all we have to start 
with. This is too "soft" even for a simulation, so asecond 
level of abstraction - which we shall call "manageable" 
theory - must be made. Our very general theory of the 
real world is narrowed in scope, the least certain ele- 
ments are dropped, irrationality is set aside, only rela- 
tionships which demonstrate logical connections are 
retained, and the quantitative is preferred to the quali- 
tative. This manageable theory is the foundation of the 
model, but is not the model itself. As an example, if we 
choose to study the movement of people around an urban 
center, general mobility could be narrowed to include 
only work trip patterns, excluding the less certain shop- 
ping and social mobility patterns. 
The third level of abstraction is the model itself - an 
explicit statement of all relationships in an internally 
logical framework. Those aspects that are too complex, 
defy measurement or relate in an unknown fashion are 
quite simply dropped. In the example above, for instance 
a single gravity equation might be used to describe work 
trip patterns. 
Converting this model to an appropriate language and 
form for the computing machinery available is the fourth 
abstraction from reality. Further exclusion and simplifi- 
Zoal:; 
Policies 
Formulate 
Relationships 
and Theories 
Generate and 
Test Alternatives: 
— 
+ 
cation may be required for practical programming; com- 
puting machinery often imposes computational constraints - 
continuous functions may have to be treated as discrete, 
data rounded off to satisfy storage limitations, and so 
Forth. 
After all this abstraction, one may well ask, "What does 
the computer output mean in terms of the real world? How 
valid is its forecast?" And since a simulation model is 
established as an experimental laboratory, "How much 
faith can I put in the experiments?" In reply we can say 
that the urban model must: 
- be internally consistent 
- contain no logical errors 
- produce reasonably accurate forecasts. 
Beyond this, one can check the soundness of the reasoning 
on which the computer program is based by going through 
the four stages of abstraction in reverse order. The simu- 
lation program is compared successively with each of the 
lower levels of abstraction in order to identify and explain 
differences. 
A final test of the model and its computer program can be 
carried out by inserting artificial data and checking that 
t+he pattern of output is consistent with the model’ s theory 
Models for Urban Planning 
The use of such analytic techniques for urban planning is 
not without challenge. Föäur arguments are put forth: 
- urban phenomena are too complex and disorderly for 
reduction to systematic models 
the indeterminacy of human behavior makes social pre- 
diction impossible 
- the nature of technological change is unpredictable 
although the fact of such innovation is certain. 
the rate of social and economic change is accelerating 
varticularly in highly developed countries 
Figure 1: Flow Diagram 
of the Planning Process 
Choose the 
"Best" Plan 
Implement 
Real World 
Project 
Trends 
Evaluate 
Measure 
Results 
ARCH+ 2 (1969) H.&
	        

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.